Overhaul

I guess an overhaul is a sane thing, even though it feels otherwise.

So, it didn't go so well. Unification, in itself, isn't a tricky process, but implementing it showed holes in the plot, things I didn't expect. I had to reshape the semantics a bit, add a few programmatic tools here and there, and in the end I'm still not satisfied with the results so far.

I’m planning a complete overhaul of Hup. There are several issues, in many domains, I want to tackle:

• The syntax is irregular, so I’ll change it to something closer to LISP, except it’s a(b c d) instead of (a b c d),
• I can’t refer to tables through variables, I want to make it possible,
• There’s no macros, I want macros + what I call “brushes”, that is, macros on steroids for highly redundant graphs,
• There’s no access to low-level stuff, I want a way to do simple things like manipulating strings for instance,
• It lacks some high-level specialized functionalities through libraries, namely 3D and ANN,
• It’s not very fast, I want to implement it in another language.

Last but not least, this time I want to draw a map of the system, get it right, step by step. A partner or a team would be a dreamy cherry on the cake.

Is C the only path when you're being serious about something? I have to find out.